At the speed of the web resource greatly affect loadable external objects. The most common of them - the image. In this connection it becomes actual size reduction loadable computer graphics (images) to reduce the overall size of the web page. Among the least popular compression algorithms algorithms for lossless coding, because of their low efficiency. However, for a certain class of images through the use of its specific features, it is possible to obtain a much higher compression ratio in lossless coding. The authors proposed a modification of the RLE algorithm, which allows to increase the compression ratio of the images with large areas of solid color.

Keywords: RLE compression algorithm, the compression ratio, data processing, evaluation of the effectiveness of the compression algorithm, the length of the sequence

Modern CAD systems now allow automate most of the stages of design and technological preparation of production . But there is a number of routine procedures. This article is about the details classification operation analysys .
 A review of existing solutions has shown that there are two ways to classify the details. The first - a “manual” characteristics of the classifier, assignment. This is the major method. The second is the expert system "Klassifikator" from the AsconCo., which is based only on the verbal description of the parts and the user selected one of parts from the existing. So that the function of user is to provide the recognition of images patterns.  These methods of pattern recognition  systems are successfully solve complex image recognition task . Analysis of methods of recognition led to the selection of the neural network method for classification of items.
We want you to use expert system consists of a Hamming neural network emulator and the semantic model of knowledge representation with direct output to solve the problem of classification of parts.
This combination of neuro-cybernetics and information approach to the creation of an expert system can either automatically determine a part characteristic by its image , or in a semi-automatic mode, prompting the missing information from user.  

Keywords: expert system, pattern recognition, computer-aided design, neural network classification characteristic details.